ON THE COLOURS OF THICK PLATES. 175
14. All the preceding conclusions agree perfectly with experiment, so far as qualitative results are concerned; for I have not taken any measures. The change in the direction of curvature, which I had not noticed till it was pointed out by the formula, may be readily seen, when the flame of a candle is the source of light, by holding the candle at arm's length nearly in front of the eye, but a little to the right, then drawing it back beside the eye, and finally holding it at arm's length behind the head, and as nearly in a line drawn through the eye perpendicular to the mirror as the shadow of the head will allow.
When the candle is held near the eye, a portion only of the circles can be seen; the circles are in fact reduced to circular arcs, and these arcs may even become perfectly straight. But when the candle is placed at a good distance from the eye, suppose half-way between the eye and the mirror, and a small piece of card is placed as a screen in front of the flame to keep off the glare of the direct light, the circles, or at least several of them, may be seen complete, except that it must be left to the imagination to fill them up where they are hid by the screen. lu this way the manner in which the rings open out from the centre of the circles may be observed, though not for the first ring or two, which open out while the centre is hid by the screen. Instead of a candle with a screen, it is better to use the image of the sun in the focus of a small convex or concave mirror.
15. The conclusion deduced from theory which was mentioned in Art. 13 cannot, of course, be compared with experiment directly. But the experiment may be successfully performed by substituting for either the luminous point or the eye a virtual image. Using for the luminous point the imago of the sun in the focus of a small concave mirror, at the distance of sonic feet in front of a plane mirror of which the surface had been prepared with milk and water, I placed a-piece of plate glass between the mirrors, inclined at an angle of about 45°. The greater part of the light coming from the image of the sun was transmitted through the plate of glass; and on returning from the large mirror a portion of this light was reflected sideways, so that the rings could be seen by reflexion in the plate of glass without obstructing the incident light. The system of rings thus seen was very beautiful, and there was no direct light glaring in the eye, and yet no screen to hide